焊缝超声波探伤中缺陷智能识别研究

Intelligent Recognition of Weld Defects in Ultrasonic Test

分析超声波A扫探伤过程中焊缝缺陷波动态波形特点,截取数个能反映动态波形特性的波形,并提取该系列波形的特征参数。根据特征参数的变化规律,利用相对变率关联度分析法,结合探头探测方式提出了一种对焊缝缺陷进行智能识别的方法。实验分析表明:相对变率法的引用能为缺陷类型提供有效的智能识别。

Through capturing several waveforms to reflect dynamic waveform characteristics and extracting their characteristic parameters,the weld defect’s dynamic waveform characteristics were analyzed in ultrasonic A-scan testing process.Basing on the change rules of characteristic parameters and the relative variability relational degree analysis as well as the probe detection method,an intelligent recognition method for weld defects was proposed.The experimental study on it indicates that application of relative variability relational degree analysis can provide an intelligent recognition of defect types.